In [ ]:
# import libraries
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.model_selection import GridSearchCV
#to display all the graph in the workbook 
%matplotlib inline
sns.set_style("whitegrid",{'grid.linestyle': '--'})
import warnings
warnings.filterwarnings("ignore")
In [ ]:
# dataset
data='SeoulBikeData.csv'
In [ ]:
# dataframe
df = pd.read_csv(data, encoding= 'unicode_escape')
In [ ]:
# Check 1st Five rows
df.head()
Out[ ]:
Date Rented Bike Count Hour Temperature(°C) Humidity(%) Wind speed (m/s) Visibility (10m) Dew point temperature(°C) Solar Radiation (MJ/m2) Rainfall(mm) Snowfall (cm) Seasons Holiday Functioning Day
0 01/12/2017 254 0 -5.2 37 2.2 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes
1 01/12/2017 204 1 -5.5 38 0.8 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes
2 01/12/2017 173 2 -6.0 39 1.0 2000 -17.7 0.0 0.0 0.0 Winter No Holiday Yes
3 01/12/2017 107 3 -6.2 40 0.9 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes
4 01/12/2017 78 4 -6.0 36 2.3 2000 -18.6 0.0 0.0 0.0 Winter No Holiday Yes
In [ ]:
# Check last 5 rows
df.tail(5)
Out[ ]:
Date Rented Bike Count Hour Temperature(°C) Humidity(%) Wind speed (m/s) Visibility (10m) Dew point temperature(°C) Solar Radiation (MJ/m2) Rainfall(mm) Snowfall (cm) Seasons Holiday Functioning Day
8755 30/11/2018 1003 19 4.2 34 2.6 1894 -10.3 0.0 0.0 0.0 Autumn No Holiday Yes
8756 30/11/2018 764 20 3.4 37 2.3 2000 -9.9 0.0 0.0 0.0 Autumn No Holiday Yes
8757 30/11/2018 694 21 2.6 39 0.3 1968 -9.9 0.0 0.0 0.0 Autumn No Holiday Yes
8758 30/11/2018 712 22 2.1 41 1.0 1859 -9.8 0.0 0.0 0.0 Autumn No Holiday Yes
8759 30/11/2018 584 23 1.9 43 1.3 1909 -9.3 0.0 0.0 0.0 Autumn No Holiday Yes
In [ ]:
# Check Total rows and Columns in dataset.
# Rows= 8760
# Columns=14
print(f' The shape of dataset is {(df.shape)} \n Total Rows are : {(df.shape)[0]}\n Total Columns are : {(df.shape)[1]}')

 The shape of dataset is (8760, 14) 
 Total Rows are : 8760
 Total Columns are : 14
In [ ]:
# Check the info of dataset to know the data types of columns or variable
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8760 entries, 0 to 8759
Data columns (total 14 columns):
 #   Column                     Non-Null Count  Dtype  
---  ------                     --------------  -----  
 0   Date                       8760 non-null   object 
 1   Rented Bike Count          8760 non-null   int64  
 2   Hour                       8760 non-null   int64  
 3   Temperature(°C)            8760 non-null   float64
 4   Humidity(%)                8760 non-null   int64  
 5   Wind speed (m/s)           8760 non-null   float64
 6   Visibility (10m)           8760 non-null   int64  
 7   Dew point temperature(°C)  8760 non-null   float64
 8   Solar Radiation (MJ/m2)    8760 non-null   float64
 9   Rainfall(mm)               8760 non-null   float64
 10  Snowfall (cm)              8760 non-null   float64
 11  Seasons                    8760 non-null   object 
 12  Holiday                    8760 non-null   object 
 13  Functioning Day            8760 non-null   object 
dtypes: float64(6), int64(4), object(4)
memory usage: 958.2+ KB
In [ ]:
# Check columns
df.columns
Out[ ]:
Index(['Date', 'Rented Bike Count', 'Hour', 'Temperature(°C)', 'Humidity(%)',
       'Wind speed (m/s)', 'Visibility (10m)', 'Dew point temperature(°C)',
       'Solar Radiation (MJ/m2)', 'Rainfall(mm)', 'Snowfall (cm)', 'Seasons',
       'Holiday', 'Functioning Day'],
      dtype='object')
In [ ]:
#Since the variables having units with name, renaming columns for better variable analysis.
df.rename(columns={'Rented Bike Count':'Rented_Bike_Count','Temperature(°C)':'Temperature','Humidity(%)':'Humidity','Wind speed (m/s)':'Wind_speed',
                       'Visibility (10m)':'Visibility','Dew point temperature(°C)':'Dew_point_temperature', 'Solar Radiation (MJ/m2)':'Solar_Radiation',
                        'Rainfall(mm)':'Rainfall','Snowfall (cm)':'Snowfall','Functioning Day':'Functioning_Day'},inplace=True)
In [ ]:
# Count missing and non-missing values in each column
missing_values_count = df.isnull().sum()
non_missing_values_count = df.notnull().sum()

# Create stacked bar chart
plt.figure(figsize=(10, 6))
plt.bar(missing_values_count.index, missing_values_count.values, label='Missing')
plt.bar(non_missing_values_count.index, non_missing_values_count.values, bottom=missing_values_count.values, label='Non-missing')
plt.xticks(rotation=90)
plt.xlabel("Column name", size=14, weight="bold")
plt.ylabel("Count", size=14, weight="bold")
plt.title("Missing and non-missing values by column", fontweight="bold", size=17)
plt.legend()
plt.show()
In [ ]:
# checking Duplicate rows in our BikeData.
duplicates=df.duplicated().sum()
print(f"We have {duplicates} duplicate rows in our Bike Data.")
# No duplicate rows found

We have 0 duplicate rows in our Bike Data.
In [ ]:
df.isna().sum()
Out[ ]:
Date                     0
Rented_Bike_Count        0
Hour                     0
Temperature              0
Humidity                 0
Wind_speed               0
Visibility               0
Dew_point_temperature    0
Solar_Radiation          0
Rainfall                 0
Snowfall                 0
Seasons                  0
Holiday                  0
Functioning_Day          0
dtype: int64
In [ ]:
# Data Description
df.describe().transpose()
Out[ ]:
count mean std min 25% 50% 75% max
Rented_Bike_Count 8760.0 704.602055 644.997468 0.0 191.00 504.50 1065.25 3556.00
Hour 8760.0 11.500000 6.922582 0.0 5.75 11.50 17.25 23.00
Temperature 8760.0 12.882922 11.944825 -17.8 3.50 13.70 22.50 39.40
Humidity 8760.0 58.226256 20.362413 0.0 42.00 57.00 74.00 98.00
Wind_speed 8760.0 1.724909 1.036300 0.0 0.90 1.50 2.30 7.40
Visibility 8760.0 1436.825799 608.298712 27.0 940.00 1698.00 2000.00 2000.00
Dew_point_temperature 8760.0 4.073813 13.060369 -30.6 -4.70 5.10 14.80 27.20
Solar_Radiation 8760.0 0.569111 0.868746 0.0 0.00 0.01 0.93 3.52
Rainfall 8760.0 0.148687 1.128193 0.0 0.00 0.00 0.00 35.00
Snowfall 8760.0 0.075068 0.436746 0.0 0.00 0.00 0.00 8.80
In [ ]:
# extracting day,month, day of week and weekdays/weekend from date column

df['Date']=pd.to_datetime(df['Date'])
df['month'] = df['Date'].apply(lambda x : x.month)
df['day_of_week'] = df['Date'].dt.day_name()
df['weekdays_weekend']=df['day_of_week'].apply(lambda x : 1 if x=='Saturday' or x=='Sunday' else 0 )
df=df.drop(columns=['Date','day_of_week'],axis=1)
In [ ]:
#checking correlation using heatmap

plt.figure(figsize=(15,10))
sns.heatmap(df.corr(),cmap='cividis',annot=True)
Out[ ]:
<Axes: >
In [ ]:
# Multicollinearity

from statsmodels.stats.outliers_influence import variance_inflation_factor
def calc_vif(X):
 
   # Calculating VIF
   vif = pd.DataFrame()
   vif["variables"] = X.columns
   vif["VIF"] = [variance_inflation_factor(X.values, i) for i in range(X.shape[1])]
 
   return(vif)
In [ ]:
calc_vif(df[[i for i in df.describe().columns if i not in ['Rented_Bike_Count','Dew_point_temperature'] ]])
Out[ ]:
variables VIF
0 Hour 3.961874
1 Temperature 3.236809
2 Humidity 6.114153
3 Wind_speed 4.616621
4 Visibility 5.404128
5 Solar_Radiation 2.272068
6 Rainfall 1.081252
7 Snowfall 1.125304
8 month 4.580307
9 weekdays_weekend 1.399708
In [ ]:
#visualising distribution

sns.distplot(df['Rented_Bike_Count'])
Out[ ]:
<Axes: xlabel='Rented_Bike_Count', ylabel='Density'>
In [ ]:
# square_root transformation

plt.figure(figsize=(10,8))
sns.distplot(np.sqrt(df['Rented_Bike_Count']))
Out[ ]:
<Axes: xlabel='Rented_Bike_Count', ylabel='Density'>
In [ ]:
# Numeric Features

numeric_features= df.select_dtypes(exclude='object')
In [ ]:
# Numeric Features info
numeric_features.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8760 entries, 0 to 8759
Data columns (total 12 columns):
 #   Column                 Non-Null Count  Dtype  
---  ------                 --------------  -----  
 0   Rented_Bike_Count      8760 non-null   int64  
 1   Hour                   8760 non-null   int64  
 2   Temperature            8760 non-null   float64
 3   Humidity               8760 non-null   int64  
 4   Wind_speed             8760 non-null   float64
 5   Visibility             8760 non-null   int64  
 6   Dew_point_temperature  8760 non-null   float64
 7   Solar_Radiation        8760 non-null   float64
 8   Rainfall               8760 non-null   float64
 9   Snowfall               8760 non-null   float64
 10  month                  8760 non-null   int64  
 11  weekdays_weekend       8760 non-null   int64  
dtypes: float64(6), int64(6)
memory usage: 821.4 KB
In [ ]:
# describing numeric feature

numeric_features.describe().transpose()
Out[ ]:
count mean std min 25% 50% 75% max
Rented_Bike_Count 8760.0 704.602055 644.997468 0.0 191.00 504.50 1065.25 3556.00
Hour 8760.0 11.500000 6.922582 0.0 5.75 11.50 17.25 23.00
Temperature 8760.0 12.882922 11.944825 -17.8 3.50 13.70 22.50 39.40
Humidity 8760.0 58.226256 20.362413 0.0 42.00 57.00 74.00 98.00
Wind_speed 8760.0 1.724909 1.036300 0.0 0.90 1.50 2.30 7.40
Visibility 8760.0 1436.825799 608.298712 27.0 940.00 1698.00 2000.00 2000.00
Dew_point_temperature 8760.0 4.073813 13.060369 -30.6 -4.70 5.10 14.80 27.20
Solar_Radiation 8760.0 0.569111 0.868746 0.0 0.00 0.01 0.93 3.52
Rainfall 8760.0 0.148687 1.128193 0.0 0.00 0.00 0.00 35.00
Snowfall 8760.0 0.075068 0.436746 0.0 0.00 0.00 0.00 8.80
month 8760.0 6.526027 3.448048 1.0 4.00 7.00 10.00 12.00
weekdays_weekend 8760.0 0.290411 0.453978 0.0 0.00 0.00 1.00 1.00
In [ ]:
#plotting histogram

for col in numeric_features[:]:
  sns.histplot(df[col])
  plt.axvline(df[col].mean(), color='magenta', linestyle='dashed', linewidth=2)
  plt.axvline(df[col].median(), color='cyan', linestyle='dashed', linewidth=2)   
  plt.show()
In [ ]:
# ploting Regression plot of each columns of dataset v/s rented bike count columns

for col in numeric_features[:]:
  if col == 'Rented_Bike_Count':
    pass
  else:
    sns.regplot(x=df[col],y=df["Rented_Bike_Count"],line_kws={"color": "red"})
  
  plt.show()
In [ ]:
#Extracting categorical features
categorical_features= df.select_dtypes(include='object')
In [ ]:
categorical_features
Out[ ]:
Seasons Holiday Functioning_Day
0 Winter No Holiday Yes
1 Winter No Holiday Yes
2 Winter No Holiday Yes
3 Winter No Holiday Yes
4 Winter No Holiday Yes
... ... ... ...
8755 Autumn No Holiday Yes
8756 Autumn No Holiday Yes
8757 Autumn No Holiday Yes
8758 Autumn No Holiday Yes
8759 Autumn No Holiday Yes

8760 rows × 3 columns

In [ ]:
#ploting Boxen plot to visualize and trying to get information from plot
for col in categorical_features:
    plt.figure(figsize=(10,8))
    sns.boxenplot(x=df[col], y=df["Rented_Bike_Count"])
    plt.show()
In [ ]:
#checking counts of functioning day
df['Functioning_Day'].value_counts()
Out[ ]:
Yes    8465
No      295
Name: Functioning_Day, dtype: int64
In [ ]:
#ploting cat plot for more info
sns.catplot(x='Seasons',y='Rented_Bike_Count',data=df)
Out[ ]:
<seaborn.axisgrid.FacetGrid at 0x7fa4809aeb90>
In [ ]:
#ploting line graph
# group by Hrs and get average Bikes rented, and precent change
avg_rent_hrs = df.groupby('Hour')['Rented_Bike_Count'].mean()

# plot average rent over time(hrs)
plt.figure(figsize=(20,4))
a=avg_rent_hrs.plot(legend=True,marker='o',title="Average Bikes Rented Per Hr")
a.set_xticks(range(len(avg_rent_hrs)));
a.set_xticklabels(avg_rent_hrs.index.tolist(), rotation=85);
In [ ]:
df.head()
Out[ ]:
Rented_Bike_Count Hour Temperature Humidity Wind_speed Visibility Dew_point_temperature Solar_Radiation Rainfall Snowfall Seasons Holiday Functioning_Day month weekdays_weekend
0 254 0 -5.2 37 2.2 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
1 204 1 -5.5 38 0.8 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
2 173 2 -6.0 39 1.0 2000 -17.7 0.0 0.0 0.0 Winter No Holiday Yes 1 0
3 107 3 -6.2 40 0.9 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
4 78 4 -6.0 36 2.3 2000 -18.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
In [ ]:
dff=df.copy()
In [ ]:
# Import the necessary libraries
from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler


# Preprocess the data
df1 = dff.dropna() # Remove missing values
df1 = dff.drop(columns=['Holiday']) # Remove irrelevant columns
df1['Seasons'] = df1['Seasons'].map({'Winter': 0, 'Spring': 1, 'Summer': 2, 'Autumn': 3}) # Convert categorical variable to numerical

# Select the features
X = dff[['Rented_Bike_Count', 'Hour', 'Temperature', 'Humidity', 'Wind_speed', 'Visibility', 'Dew_point_temperature', 'Solar_Radiation', 'Rainfall']].values

# Scale the features
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

# Determine the optimal number of clusters using the elbow method
sse = []
for k in range(1, 11):
    kmeans = KMeans(n_clusters=k, random_state=42)
    kmeans.fit(X_scaled)
    sse.append(kmeans.inertia_)
plt.plot(range(1, 11), sse)
plt.xlabel('Number of clusters')
plt.ylabel('SSE')
plt.show()

# Perform clustering using the optimal number of clusters
kmeans = KMeans(n_clusters=4, random_state=42)
kmeans.fit(X_scaled)
labels = kmeans.labels_

# Visualize the clusters
dff['Cluster'] = labels
dff.plot(kind='scatter', x='Rented_Bike_Count', y='Temperature', c='Cluster', cmap='viridis')
plt.xlabel('Rented_Bike_Count')
plt.ylabel('Temperature')
plt.show()
In [ ]:
df.head()
Out[ ]:
Rented_Bike_Count Hour Temperature Humidity Wind_speed Visibility Dew_point_temperature Solar_Radiation Rainfall Snowfall Seasons Holiday Functioning_Day month weekdays_weekend
0 254 0 -5.2 37 2.2 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
1 204 1 -5.5 38 0.8 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
2 173 2 -6.0 39 1.0 2000 -17.7 0.0 0.0 0.0 Winter No Holiday Yes 1 0
3 107 3 -6.2 40 0.9 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
4 78 4 -6.0 36 2.3 2000 -18.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
In [ ]:
# make copy of dataset

df2=df.copy()
In [ ]:
df2.head()
Out[ ]:
Rented_Bike_Count Hour Temperature Humidity Wind_speed Visibility Dew_point_temperature Solar_Radiation Rainfall Snowfall Seasons Holiday Functioning_Day month weekdays_weekend
0 254 0 -5.2 37 2.2 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
1 204 1 -5.5 38 0.8 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
2 173 2 -6.0 39 1.0 2000 -17.7 0.0 0.0 0.0 Winter No Holiday Yes 1 0
3 107 3 -6.2 40 0.9 2000 -17.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
4 78 4 -6.0 36 2.3 2000 -18.6 0.0 0.0 0.0 Winter No Holiday Yes 1 0
In [ ]:
#Multicollinearity
from statsmodels.stats.outliers_influence import variance_inflation_factor
def Calculate_vif(X):

    # Calculating VIF
    vif = pd.DataFrame()
    vif["variables"] = X.columns
    vif["VIF"] = [variance_inflation_factor(X.values, i) for i in range(X.shape[1])]

    return(vif)
In [ ]:
Calculate_vif(df2[[i for i in df2.describe().columns if i not in ['Rented_Bike_Count']]])
Out[ ]:
variables VIF
0 Hour 4.424147
1 Temperature 34.331857
2 Humidity 6.973710
3 Wind_speed 4.835871
4 Visibility 9.309721
5 Dew_point_temperature 17.758961
6 Solar_Radiation 2.881777
7 Rainfall 1.082987
8 Snowfall 1.125842
9 month 4.690212
10 weekdays_weekend 1.416551
In [ ]:
Calculate_vif(df2[[i for i in df2.describe().columns if i not in ['Rented_Bike_Count','Humidity']]])
Out[ ]:
variables VIF
0 Hour 4.261971
1 Temperature 32.943543
2 Wind_speed 4.359571
3 Visibility 8.411644
4 Dew_point_temperature 15.570048
5 Solar_Radiation 2.791031
6 Rainfall 1.056313
7 Snowfall 1.090246
8 month 3.741733
9 weekdays_weekend 1.379568
In [ ]:
Calculate_vif(df2[[i for i in df2.describe().columns if i not in ['Rented_Bike_Count','Humidity','Visibility']]])
Out[ ]:
variables VIF
0 Hour 4.261758
1 Temperature 20.568971
2 Wind_speed 4.160796
3 Dew_point_temperature 9.867059
4 Solar_Radiation 2.485410
5 Rainfall 1.055283
6 Snowfall 1.090139
7 month 3.294112
8 weekdays_weekend 1.372579
In [ ]:
# As per above vif calculation dropping humidity and visibility columns.
df2.drop(['Humidity','Visibility'],inplace=True,axis=1)
In [ ]:
# checking categoiries in season column.
df2['Seasons'].value_counts()
Out[ ]:
Spring    2208
Summer    2208
Autumn    2184
Winter    2160
Name: Seasons, dtype: int64
In [ ]:
# Createing dummy variables
df2=pd.get_dummies(df2,columns=['Seasons'],prefix='Seasons',drop_first=True)
In [ ]:
# checking categoiries in holiday columns.
df2['Holiday'].value_counts()
Out[ ]:
No Holiday    8328
Holiday        432
Name: Holiday, dtype: int64
In [ ]:
# Labeling for holiday=1 and no holiday=0
df2['Holiday']=df2['Holiday'].map({'No Holiday':0, 'Holiday':1})
In [ ]:
# # Labeling for Yes=1 and no No=0
df2['Functioning_Day']=df2['Functioning_Day'].map({'Yes':1, 'No':0})
In [ ]:
# checking dummy and labled columns
df2.head(3)
Out[ ]:
Rented_Bike_Count Hour Temperature Wind_speed Dew_point_temperature Solar_Radiation Rainfall Snowfall Holiday Functioning_Day month weekdays_weekend Seasons_Spring Seasons_Summer Seasons_Winter
0 254 0 -5.2 2.2 -17.6 0.0 0.0 0.0 0 1 1 0 0 0 1
1 204 1 -5.5 0.8 -17.6 0.0 0.0 0.0 0 1 1 0 0 0 1
2 173 2 -6.0 1.0 -17.7 0.0 0.0 0.0 0 1 1 0 0 0 1
In [ ]:
# final info before going for modeling
df2.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8760 entries, 0 to 8759
Data columns (total 15 columns):
 #   Column                 Non-Null Count  Dtype  
---  ------                 --------------  -----  
 0   Rented_Bike_Count      8760 non-null   int64  
 1   Hour                   8760 non-null   int64  
 2   Temperature            8760 non-null   float64
 3   Wind_speed             8760 non-null   float64
 4   Dew_point_temperature  8760 non-null   float64
 5   Solar_Radiation        8760 non-null   float64
 6   Rainfall               8760 non-null   float64
 7   Snowfall               8760 non-null   float64
 8   Holiday                8760 non-null   int64  
 9   Functioning_Day        8760 non-null   int64  
 10  month                  8760 non-null   int64  
 11  weekdays_weekend       8760 non-null   int64  
 12  Seasons_Spring         8760 non-null   uint8  
 13  Seasons_Summer         8760 non-null   uint8  
 14  Seasons_Winter         8760 non-null   uint8  
dtypes: float64(6), int64(6), uint8(3)
memory usage: 847.0 KB
In [ ]:
df2.head()
Out[ ]:
Rented_Bike_Count Hour Temperature Wind_speed Dew_point_temperature Solar_Radiation Rainfall Snowfall Holiday Functioning_Day month weekdays_weekend Seasons_Spring Seasons_Summer Seasons_Winter
0 254 0 -5.2 2.2 -17.6 0.0 0.0 0.0 0 1 1 0 0 0 1
1 204 1 -5.5 0.8 -17.6 0.0 0.0 0.0 0 1 1 0 0 0 1
2 173 2 -6.0 1.0 -17.7 0.0 0.0 0.0 0 1 1 0 0 0 1
3 107 3 -6.2 0.9 -17.6 0.0 0.0 0.0 0 1 1 0 0 0 1
4 78 4 -6.0 2.3 -18.6 0.0 0.0 0.0 0 1 1 0 0 0 1
In [ ]:
# Fuzzy C means
In [ ]:
import pandas as pd
import numpy as np
from sklearn.cluster import KMeans
from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import silhouette_score
from sklearn.metrics.pairwise import euclidean_distances


# Preprocess the data
X = df2.iloc[:, 3:].values
scaler = StandardScaler()
X = scaler.fit_transform(X)

# Determine the optimal number of clusters
silhouette_scores = []
for k in range(2, 11):
    kmeans = KMeans(n_clusters=k, random_state=42)
    kmeans.fit(X)
    labels = kmeans.predict(X)
    silhouette_scores.append(silhouette_score(X, labels))

# Choose the number of clusters with the highest silhouette score
n_clusters = np.argmax(silhouette_scores) + 2
n_clusters_reduced = 4 # set the reduced number of clusters

# Perform Fuzzy C-Means clustering
def cmeans(X, n_clusters, m=2, max_iter=100, tol=1e-4):
    # Initialize the cluster centroids and membership matrix
    C = X[np.random.choice(len(X), n_clusters, replace=False)]
    U = np.random.dirichlet(np.ones(n_clusters), len(X))

    # Iterate until convergence or max_iter is reached
    for i in range(max_iter):
        # Update the cluster centroids
        C_old = C.copy()
        for j in range(n_clusters):
            C[j] = np.sum(U[:, j].reshape(-1, 1) ** m * X, axis=0) / np.sum(U[:, j] ** m)

        # Update the membership matrix
        dist = euclidean_distances(X, C)
        U = 1 / np.power(dist, 2 / (m - 1))
        U /= np.sum(U, axis=1).reshape(-1, 1)

        # Check for convergence
        if np.linalg.norm(C - C_old) < tol:
            break

    return C, U

C, U = cmeans(X, n_clusters_reduced, m=2, max_iter=100, tol=1e-4)

# Apply PCA to reduce the dimensionality of the data
pca = PCA(n_components=2)
X_pca = pca.fit_transform(X)

# Plot the results
import matplotlib.pyplot as plt

fig, ax = plt.subplots()
for i in range(n_clusters_reduced):
    ax.scatter(X_pca[U[:, i] == np.max(U, axis=1), 0], X_pca[U[:, i] == np.max(U, axis=1), 1], label=f"Cluster {i + 1}")
ax.scatter(C[:, 0], C[:, 1], marker="x", color="black", label="Centroids")
ax.legend()
plt.show()
In [ ]:
In [ ]:
!pip install factor_analyzer

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting factor_analyzer
  Downloading factor_analyzer-0.4.1.tar.gz (41 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 41.8/41.8 kB 5.3 MB/s eta 0:00:00
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
  Preparing metadata (pyproject.toml) ... done
Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from factor_analyzer) (1.10.1)
Requirement already satisfied: scikit-learn in /usr/local/lib/python3.10/dist-packages (from factor_analyzer) (1.2.2)
Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (from factor_analyzer) (1.5.3)
Collecting pre-commit
  Downloading pre_commit-3.3.1-py2.py3-none-any.whl (202 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 202.5/202.5 kB 22.0 MB/s eta 0:00:00
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from factor_analyzer) (1.22.4)
Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.10/dist-packages (from pandas->factor_analyzer) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas->factor_analyzer) (2022.7.1)
Collecting nodeenv>=0.11.1
  Downloading nodeenv-1.7.0-py2.py3-none-any.whl (21 kB)
Collecting cfgv>=2.0.0
  Downloading cfgv-3.3.1-py2.py3-none-any.whl (7.3 kB)
Collecting virtualenv>=20.10.0
  Downloading virtualenv-20.23.0-py3-none-any.whl (3.3 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 3.3/3.3 MB 88.2 MB/s eta 0:00:00
Collecting identify>=1.0.0
  Downloading identify-2.5.24-py2.py3-none-any.whl (98 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 98.8/98.8 kB 13.0 MB/s eta 0:00:00
Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from pre-commit->factor_analyzer) (6.0)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->factor_analyzer) (3.1.0)
Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->factor_analyzer) (1.2.0)
Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from nodeenv>=0.11.1->pre-commit->factor_analyzer) (67.7.2)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.1->pandas->factor_analyzer) (1.16.0)
Requirement already satisfied: platformdirs<4,>=3.2 in /usr/local/lib/python3.10/dist-packages (from virtualenv>=20.10.0->pre-commit->factor_analyzer) (3.3.0)
Collecting distlib<1,>=0.3.6
  Downloading distlib-0.3.6-py2.py3-none-any.whl (468 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 468.5/468.5 kB 42.6 MB/s eta 0:00:00
Requirement already satisfied: filelock<4,>=3.11 in /usr/local/lib/python3.10/dist-packages (from virtualenv>=20.10.0->pre-commit->factor_analyzer) (3.12.0)
Building wheels for collected packages: factor_analyzer
  Building wheel for factor_analyzer (pyproject.toml) ... done
  Created wheel for factor_analyzer: filename=factor_analyzer-0.4.1-py2.py3-none-any.whl size=42034 sha256=66d9ea23e0f1d3a9d515c30fb34b7cd02a6129430e80312c5d51e96e1d20bdeb
  Stored in directory: /root/.cache/pip/wheels/c5/94/da/41abe415f64706710726291086a814dd8b9e0dab1c491ef6ed
Successfully built factor_analyzer
Installing collected packages: distlib, virtualenv, nodeenv, identify, cfgv, pre-commit, factor_analyzer
Successfully installed cfgv-3.3.1 distlib-0.3.6 factor_analyzer-0.4.1 identify-2.5.24 nodeenv-1.7.0 pre-commit-3.3.1 virtualenv-20.23.0
In [ ]:
# FACTOR ANALYSIS
import numpy as np
from factor_analyzer import FactorAnalyzer

# initialize factor analyzer object
fa = FactorAnalyzer()

# specify number of factors to extract
num_factors = 3

# fit factor analyzer to data and extract factors
fa.fit(df2)
loadings = fa.loadings_

# print factor loadings
print("Factor loadings:\n", loadings)

Factor loadings:
 [[ 4.79375576e-01  8.81828560e-02  4.62245925e-01]
 [ 5.57754992e-02  1.59709290e-02  5.22357534e-01]
 [ 9.43733969e-01  2.12086675e-01  5.44288695e-02]
 [-8.43681076e-02 -1.04116114e-01  5.26634364e-01]
 [ 8.99110987e-01  2.20849470e-01 -2.55159144e-01]
 [ 2.31901372e-01 -3.90093989e-02  3.93868358e-01]
 [ 6.65212954e-02  8.11463963e-05 -1.33770210e-01]
 [-2.37442074e-01  7.00862928e-02 -1.05868017e-02]
 [-9.63328888e-02  1.57476765e-02  1.96246601e-02]
 [-2.38683118e-02  1.99177702e-02  1.18474947e-01]
 [ 1.79635791e-02  2.17098876e-01 -2.62300421e-02]
 [-1.90797176e-02  2.45023719e-03 -1.25789454e-02]
 [ 3.59561049e-01 -9.90931037e-01 -1.28188241e-01]
 [ 4.97358572e-01  4.76529827e-01 -8.31071661e-03]
 [-8.50110194e-01  1.57483502e-01  1.72065805e-01]]
In [ ]:
ev, v = fa.get_eigenvalues()
ev
Out[ ]:
array([3.52693437, 1.74085752, 1.4415795 , 1.13234251, 1.05769033,
       1.02096331, 0.99790337, 0.95519985, 0.88412224, 0.82750147,
       0.57553934, 0.39000094, 0.24093575, 0.17204929, 0.03638021])
In [ ]:
sorted(ev)
Out[ ]:
[0.03638021011887888,
 0.17204928821536944,
 0.2409357546017857,
 0.3900009367621154,
 0.5755393422591972,
 0.8275014740066442,
 0.8841222410148962,
 0.9551998473516831,
 0.9979033689630333,
 1.0209633083506582,
 1.0576903263597195,
 1.1323425105047848,
 1.441579502244821,
 1.7408575194601552,
 3.526934369786267]
In [ ]:
plt.scatter(range(1,df2.shape[1]+1),ev)
plt.plot(range(1,df2.shape[1]+1),ev)
plt.title('Scree Plot')
plt.xlabel('Factors')
plt.ylabel('Eigen Value')
plt.grid()
plt.show()
In [ ]:
import matplotlib.pyplot as plt
import seaborn as sns

# Create a heatmap of the factor loadings
plt.figure(figsize=(10, 8))
sns.heatmap(loadings, cmap='coolwarm', annot=True, fmt='.2f', 
            xticklabels=['Factor 1', 'Factor 2', 'Factor 3'], 
            yticklabels=df2.columns[:])
plt.title('Factor Loadings Heatmap')
plt.show()
In [ ]:
In [ ]:
#creating Dummy variable for categorical columns
dummy_categorical_feature= pd.get_dummies(categorical_features,drop_first=True)
In [ ]:
dummy_categorical_feature
Out[ ]:
Seasons_Spring Seasons_Summer Seasons_Winter Holiday_No Holiday Functioning_Day_Yes
0 0 0 1 1 1
1 0 0 1 1 1
2 0 0 1 1 1
3 0 0 1 1 1
4 0 0 1 1 1
... ... ... ... ... ...
8755 0 0 0 1 1
8756 0 0 0 1 1
8757 0 0 0 1 1
8758 0 0 0 1 1
8759 0 0 0 1 1

8760 rows × 5 columns

In [ ]:
#concating numeric columns and dummy columns and creating final df
final_df= pd.concat([dummy_categorical_feature,numeric_features],axis=1)
In [ ]:
#showing head of final df
final_df.head()
Out[ ]:
Seasons_Spring Seasons_Summer Seasons_Winter Holiday_No Holiday Functioning_Day_Yes Rented_Bike_Count Hour Temperature Humidity Wind_speed Visibility Dew_point_temperature Solar_Radiation Rainfall Snowfall month weekdays_weekend
0 0 0 1 1 1 254 0 -5.2 37 2.2 2000 -17.6 0.0 0.0 0.0 1 0
1 0 0 1 1 1 204 1 -5.5 38 0.8 2000 -17.6 0.0 0.0 0.0 1 0
2 0 0 1 1 1 173 2 -6.0 39 1.0 2000 -17.7 0.0 0.0 0.0 1 0
3 0 0 1 1 1 107 3 -6.2 40 0.9 2000 -17.6 0.0 0.0 0.0 1 0
4 0 0 1 1 1 78 4 -6.0 36 2.3 2000 -18.6 0.0 0.0 0.0 1 0
In [ ]:
#describing final DF
final_df.describe().transpose()
Out[ ]:
count mean std min 25% 50% 75% max
Seasons_Spring 8760.0 0.252055 0.434217 0.0 0.00 0.00 1.00 1.00
Seasons_Summer 8760.0 0.252055 0.434217 0.0 0.00 0.00 1.00 1.00
Seasons_Winter 8760.0 0.246575 0.431042 0.0 0.00 0.00 0.00 1.00
Holiday_No Holiday 8760.0 0.950685 0.216537 0.0 1.00 1.00 1.00 1.00
Functioning_Day_Yes 8760.0 0.966324 0.180404 0.0 1.00 1.00 1.00 1.00
Rented_Bike_Count 8760.0 704.602055 644.997468 0.0 191.00 504.50 1065.25 3556.00
Hour 8760.0 11.500000 6.922582 0.0 5.75 11.50 17.25 23.00
Temperature 8760.0 12.882922 11.944825 -17.8 3.50 13.70 22.50 39.40
Humidity 8760.0 58.226256 20.362413 0.0 42.00 57.00 74.00 98.00
Wind_speed 8760.0 1.724909 1.036300 0.0 0.90 1.50 2.30 7.40
Visibility 8760.0 1436.825799 608.298712 27.0 940.00 1698.00 2000.00 2000.00
Dew_point_temperature 8760.0 4.073813 13.060369 -30.6 -4.70 5.10 14.80 27.20
Solar_Radiation 8760.0 0.569111 0.868746 0.0 0.00 0.01 0.93 3.52
Rainfall 8760.0 0.148687 1.128193 0.0 0.00 0.00 0.00 35.00
Snowfall 8760.0 0.075068 0.436746 0.0 0.00 0.00 0.00 8.80
month 8760.0 6.526027 3.448048 1.0 4.00 7.00 10.00 12.00
weekdays_weekend 8760.0 0.290411 0.453978 0.0 0.00 0.00 1.00 1.00
In [ ]:
# ploting cat plot on some feature to get more information

feature_list=["Hour","Holiday",'Rainfall','Snowfall','weekdays_weekend']
for feature in feature_list:
  plt.figure(figsize=(10,8),dpi=200)
  sns.catplot(x=feature,y='Rented_Bike_Count',data=df)
  plt.show()

<Figure size 2000x1600 with 0 Axes>
<Figure size 2000x1600 with 0 Axes>
<Figure size 2000x1600 with 0 Axes>
<Figure size 2000x1600 with 0 Axes>
<Figure size 2000x1600 with 0 Axes>
In [ ]:
plt.figure(figsize=(7,4),dpi=100)
sns.boxplot(x='month',y='Rented_Bike_Count',data=df)
plt.show()
In [ ]:
#Functions
#Creating Function to train linear models and calculate scores
In [ ]:
#creating list of matrix to store the evaluation matrix of all model
mean_sq_error=[]
root_mean_sq_error=[]
r2_list=[]
adj_r2_list=[]
In [ ]:
# making a function to train and evaluate linear model
def train_linear_model (model,X_train,y_train,X_test,y_test):

#training the model
  model.fit(X_train,y_train)

  #predicting the values
  pred = model.predict(X_test)
  import math
  from sklearn.metrics import mean_squared_error,mean_absolute_error,r2_score

  print('\n================Evalution Matrix=========================\n')

  #finding mean_squared_error
  MSE  = mean_squared_error(y_test**2,pred**2)
  print("MSE :" , MSE)

  #finding root mean squared error
  RMSE = np.sqrt(MSE)
  print("RMSE :" ,RMSE)

  #finding the r2 score
  r2 = r2_score(y_test**2,pred**2)
  print("R2 :" ,r2)
  #finding the adjusted r2 score
  adj_r2=1-(1-r2_score(y_test**2,pred**2))*((X_test.shape[0]-1)/(X_test.shape[0]-X_test.shape[1]-1))
  print("Adjusted R2 : ",adj_r2)
  
  print('\n================Evalution Matrix=========================\n')
  

#appending metrice to list
  mean_sq_error.append(MSE)
  root_mean_sq_error.append(RMSE)
  r2_list.append(r2)
  adj_r2_list.append(adj_r2)
  # ploting the line graph of actual and predicted values
  print('\
  =========================================================\
  Evalution Graph\
  ===================================================\n')  
  plt.figure(figsize=(20,10))
  plt.plot((pred)[:100])
  plt.plot((np.array(y_test)[:100]))
  plt.legend(["Predicted","Actual"])
  plt.show()
In [ ]:
#creating function to run diffirent models
In [ ]:
def run_model(model,X_train,y_train,X_test,y_test):
    
    # Fit Model
    model.fit(X_train,y_train)
    
    # Get Metrics
    
    preds = model.predict(X_test)
    
    import math
    from sklearn.metrics import mean_squared_error
    from sklearn.metrics import mean_absolute_error
    from sklearn.metrics import r2_score
    print('\n================Evalution Matrix=========================\n')
    MSE  = mean_squared_error(y_test,preds)
    print("MSE :" , MSE)
    
    RMSE = np.sqrt(MSE)
    print("RMSE :" ,RMSE)
    r2 = r2_score(y_test,preds)
    print("R2 :" ,r2)
    adj_r2=1-(1-r2_score(y_test,preds))*((X_test.shape[0]-1)/(X_test.shape[0]-X_test.shape[1]-1))
    print("Adjusted R2 : ",adj_r2)

    train_class_preds = linear_model.predict(X_train)

    print('\n================Evalution Matrix=========================\n')

    mean_sq_error.append(MSE)
    root_mean_sq_error.append(RMSE)
    r2_list.append(r2)
    adj_r2_list.append(adj_r2)

    #ploting 
    print('\
    =========================================================\
    Evalution Graph\
    ===================================================\n') 
    plt.figure(figsize=(20,10))
    plt.plot(np.array(y_test[:100]))
    plt.plot(preds[:100])    
    plt.legend(["ACTUAL","PREDICTED"],prop={'size': 20})
    plt.show()
In [ ]:
#creating function to get feature importance
In [ ]:
def get_features_importance (optimal_model,X_train):
  '''
  shows the graph of feature importance
  '''
  features = X_train.columns
  importances = optimal_model.feature_importances_
  indices = np.argsort(importances)

  plt.figure(figsize=(15,10))
  plt.title('Feature Importance')
  plt.barh(range(len(indices)), importances[indices], color='red', align='center')
  plt.yticks(range(len(indices)), [features[i] for i in indices])
  plt.xlabel('Relative Importance')

  plt.show()
In [ ]:
#lime tablular
In [ ]:
!pip install lime

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting lime
  Downloading lime-0.2.0.1.tar.gz (275 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 275.7/275.7 kB 7.3 MB/s eta 0:00:00
  Preparing metadata (setup.py) ... done
Requirement already satisfied: matplotlib in /usr/local/lib/python3.10/dist-packages (from lime) (3.7.1)
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from lime) (1.22.4)
Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from lime) (1.10.1)
Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from lime) (4.65.0)
Requirement already satisfied: scikit-learn>=0.18 in /usr/local/lib/python3.10/dist-packages (from lime) (1.2.2)
Requirement already satisfied: scikit-image>=0.12 in /usr/local/lib/python3.10/dist-packages (from lime) (0.19.3)
Requirement already satisfied: tifffile>=2019.7.26 in /usr/local/lib/python3.10/dist-packages (from scikit-image>=0.12->lime) (2023.4.12)
Requirement already satisfied: networkx>=2.2 in /usr/local/lib/python3.10/dist-packages (from scikit-image>=0.12->lime) (3.1)
Requirement already satisfied: PyWavelets>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-image>=0.12->lime) (1.4.1)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from scikit-image>=0.12->lime) (23.1)
Requirement already satisfied: pillow!=7.1.0,!=7.1.1,!=8.3.0,>=6.1.0 in /usr/local/lib/python3.10/dist-packages (from scikit-image>=0.12->lime) (8.4.0)
Requirement already satisfied: imageio>=2.4.1 in /usr/local/lib/python3.10/dist-packages (from scikit-image>=0.12->lime) (2.25.1)
Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=0.18->lime) (1.2.0)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=0.18->lime) (3.1.0)
Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->lime) (1.0.7)
Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->lime) (3.0.9)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib->lime) (0.11.0)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->lime) (1.4.4)
Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->lime) (4.39.3)
Requirement already satisfied: python-dateutil>=2.7 in /usr/local/lib/python3.10/dist-packages (from matplotlib->lime) (2.8.2)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.7->matplotlib->lime) (1.16.0)
Building wheels for collected packages: lime
  Building wheel for lime (setup.py) ... done
  Created wheel for lime: filename=lime-0.2.0.1-py3-none-any.whl size=283859 sha256=28f8fef957efc3213254fab7d89272d45a0dd5cc0164de236e400791fefc85de
  Stored in directory: /root/.cache/pip/wheels/fd/a2/af/9ac0a1a85a27f314a06b39e1f492bee1547d52549a4606ed89
Successfully built lime
Installing collected packages: lime
Successfully installed lime-0.2.0.1
In [ ]:
# LIME SECTION

import lime
import lime.lime_tabular
def LimeTabularExplainer_table(model,instance,num_features,dependent_var):
  
  # vaariable to store feature names
  # X=df.drop(dependent_var,axis=1)
  # feature_names =list(X.columns)
  X=X_test
  feature_names =list(X_test.columns)

  explainer = lime.lime_tabular.LimeTabularExplainer(X.values,feature_names=feature_names,mode='regression')

  # selecting the instace to explain
  choosen_instance = X.reset_index(drop=True).loc[[instance]].values[0]
  exp = explainer.explain_instance(choosen_instance, model.predict, num_features=num_features)
  exp.show_in_notebook(show_table=True)
In [ ]:
#Train | Test Split for LR
In [ ]:
# Train - Spliting  data
X=final_df.drop(['Rented_Bike_Count'],axis=1)
In [ ]:
X.shape
Out[ ]:
(8760, 16)
In [ ]:
y=np.sqrt(final_df['Rented_Bike_Count'])
In [ ]:
#spliting

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=1)
In [ ]:
#scaling data
from sklearn.preprocessing import StandardScaler
std = StandardScaler()
X_train = std.fit_transform(X_train)
X_test = std.transform(X_test)
In [ ]:
#Linear Regresseion

from sklearn.linear_model import LinearRegression

#linear regression model

linear_model= LinearRegression()
In [ ]:
train_linear_model(linear_model, X_train, y_train, X_test, y_test)

================Evalution Matrix=========================

MSE : 175590.55287332056
RMSE : 419.03526447462696
R2 : 0.5729108337712394
Adjusted R2 :  0.5697661367350406

================Evalution Matrix=========================

  =========================================================  Evalution Graph  ===================================================
In [ ]:
#Regularisation
In [ ]:
#Lasso Regression
from sklearn.linear_model import Lasso
In [ ]:
# finding the best parameters for lasso by gridsearchcv
from sklearn.model_selection import GridSearchCV
lasso_model = Lasso()
#setting the parameters of the lasso model
parameters = {'alpha': [1e-15,1e-13,1e-10,1e-8,1e-5,1e-4,1e-3,1e-2,1e-1,1,5,10,20,30,40,45,50,55,60,100,0.0014]}
lasso_grid = GridSearchCV(lasso_model, parameters, scoring='neg_mean_squared_error', cv=5)
In [ ]:
# calling train_linear_model to train,fit and evalution of lasso model
train_linear_model(lasso_grid,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 175560.90711843758
RMSE : 418.9998891627987
R2 : 0.5729829411854939
Adjusted R2 :  0.569838775082856

================Evalution Matrix=========================

  =========================================================  Evalution Graph  ===================================================
In [ ]:
#Ridge Regression
from sklearn.linear_model import Ridge
ridge = Ridge()

# finding the best parameters for ridge model by gridsearchcv

parameters = {'alpha': [1e-15,1e-10,1e-8,1e-5,1e-4,1e-3,1e-2,1,5,10,20,30,35,40,45,50,55,60,100,0.5,1.5,1.6,1.7,1.8,1.9]}
ridge_grid = GridSearchCV(ridge, parameters, scoring='neg_mean_squared_error', cv=3)
In [ ]:
# calling train_linear_model to train,fit and evalution of ridge model

train_linear_model(ridge_grid,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 175248.93506569506
RMSE : 418.6274418450074
R2 : 0.5737417512792726
Adjusted R2 :  0.5706031723655443

================Evalution Matrix=========================

  =========================================================  Evalution Graph  ===================================================
In [ ]:
#ElasticNet Regression

from sklearn.linear_model import ElasticNet
elastic = ElasticNet()

# finding the best parameters for ridge model by gridsearchcv

parameters = {'alpha': [1e-15,1e-13,1e-10,1e-8,1e-5,1e-4,1e-3,1e-2,1e-1,1,5,10,20,30,40,45,50,55,60,100],'l1_ratio':[0.1,0.2,0.3,0.4,0.5]}
elastic_grid = GridSearchCV(elastic, parameters, scoring='neg_mean_squared_error',cv=5)
In [ ]:
# calling train_linear_model to train,fit and evalution of elasticnet model
train_linear_model(elastic_grid,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 175346.86749857932
RMSE : 418.74439399062925
R2 : 0.5735035500752621
Adjusted R2 :  0.5703632172640354

================Evalution Matrix=========================

  =========================================================  Evalution Graph  ===================================================
In [ ]:
#Polynomial Regression
#creating pipeline to convert x to polynomial feature and apply Linear Regression

from sklearn.pipeline import make_pipeline
#importing polynomial features
from sklearn.preprocessing import PolynomialFeatures
In [ ]:
pipe=make_pipeline(PolynomialFeatures(degree=2),LinearRegression())
In [ ]:
# calling train_linear_model to train,fit and evalution of polynomial model
train_linear_model(pipe,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 118205.13966619587
RMSE : 343.8097434137024
R2 : 0.7124894607489991
Adjusted R2 :  0.7103724940540999

================Evalution Matrix=========================

  =========================================================  Evalution Graph  ===================================================
In [ ]:
#KNN and Tree Based Method
#Train | Test split Data For Tree based methods
In [ ]:
In [ ]:
#creatin X for independent variable
X=final_df.drop(['Rented_Bike_Count'],axis=1)
In [ ]:
#creating y for dependent variable
y=final_df['Rented_Bike_Count']
In [ ]:
#spliting X and y to train and test data
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=42)
In [ ]:
#KNN Regression
from sklearn.neighbors import KNeighborsRegressor
from sklearn.model_selection import GridSearchCV
from sklearn.preprocessing import MinMaxScaler
In [ ]:
In [ ]:
#creating model
scaler = MinMaxScaler()
In [ ]:
#scaling all the feature by min max scaler
X_train_scaled=scaler.fit_transform(X_train)
X_test_scaled=scaler.transform(X_test)
In [ ]:
# creating KNN model
knn_model = KNeighborsRegressor(n_jobs=-1)

# finding the best parameters for knn_model by gridsearchcv
param_dict={'n_neighbors':[3,5,7,9,11,13,15,21,23,25]}
knn_grid= GridSearchCV(knn_model,param_grid=param_dict)

# training and evaluating the PolynomialFeatures
run_model(knn_grid,X_train_scaled,y_train,X_test_scaled,y_test)

================Evalution Matrix=========================

MSE : 83411.75920852358
RMSE : 288.81094025075225
R2 : 0.7961594359879383
Adjusted R2 :  0.7946585390600998

================Evalution Matrix=========================

    =========================================================    Evalution Graph    ===================================================
In [ ]:
In [ ]:
#Decision Tree Regression
from sklearn.tree import DecisionTreeRegressor
In [ ]:
#creating decision tree model
dt_model = DecisionTreeRegressor()
#calling run_model to train,fit and evalution of decision tree model
run_model(dt_model,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 90646.61369863014
RMSE : 301.07576072913963
R2 : 0.7784789934004401
Adjusted R2 :  0.7768479137384093

================Evalution Matrix=========================

    =========================================================    Evalution Graph    ===================================================
In [ ]:
#getting features importance of decision tree
get_features_importance(dt_model,X_train)
In [ ]:
#Random Forest
#importing Random forest
from sklearn.ensemble import RandomForestRegressor
In [ ]:
#creating model of random forest
rf_model = RandomForestRegressor()
In [ ]:
#creating param dict to check random forest with diffirent value of parameter through gridsearch
n_estimators=[60,80,100]
max_depth=[15,20]
max_leaf_nodes=[40,60,80]
params = {'n_estimators':n_estimators,'max_depth':max_depth ,'max_leaf_nodes':max_leaf_nodes}
In [ ]:
#creating rf_grid model to run rf model with gridsearch
rf_grid= GridSearchCV(rf_model,param_grid=params,verbose=0)
In [ ]:
#calling run_model to train,fit and evalution of decision tree model
run_model(rf_grid,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 62691.750320122395
RMSE : 250.3832069451192
R2 : 0.8467947221660894
Adjusted R2 :  0.8456666575340864

================Evalution Matrix=========================

    =========================================================    Evalution Graph    ===================================================
In [ ]:
#optimal rf model
optimal_rf_model=rf_grid.best_estimator_
In [ ]:
#getting and visualizing features importance  of rf model
get_features_importance(optimal_rf_model,X_train)
In [ ]:
#Gradient Boosting
#importing Gradient Boosting Regressor
from sklearn.ensemble import GradientBoostingRegressor
In [ ]:
#creating model of Gradient Boosting
gb_model=GradientBoostingRegressor()
In [ ]:
# creating param dict to check diffirent value of parameter
n_estimators=[80,100,150]
max_depth=[15,20,30]

params = {'n_estimators':n_estimators,'max_depth':max_depth }
In [ ]:
#grid search for gradient bossting
gb_grid= GridSearchCV(gb_model,param_grid=params,verbose=0)
In [ ]:
#calling run_model to train,fit and evalution of gradient boosting model
run_model(gb_grid,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 54614.730268819076
RMSE : 233.69794665084046
R2 : 0.8665332379151514
Adjusted R2 :  0.8655505098003987

================Evalution Matrix=========================

    =========================================================    Evalution Graph    ===================================================
In [ ]:
#getting best estimator of GB model given by gb_grid model
optimal_gb_model=gb_grid.best_estimator_
In [ ]:
#getting and visualizing features_importance of gradient boosting
get_features_importance(optimal_gb_model,X_train)
In [ ]:
In [ ]:
#eXtreme Gradient Boosting
#importing XGBoost Regressor
from xgboost import XGBRegressor
In [ ]:
#creating Xgboost model
xgb_model=XGBRegressor()
In [ ]:
#creating param dict for gridsearch
n_estimators=[80,100,150]
max_depth=[15,20,30]


params = {'n_estimators':n_estimators,'max_depth':max_depth }
In [ ]:
#creating xgb grid model
xgb_grid= GridSearchCV(xgb_model,param_grid=params,verbose=0)
In [ ]:
#calling run_model to train,fit and evalution of xgb_grid model
run_model(xgb_grid,X_train,y_train,X_test,y_test);

================Evalution Matrix=========================

MSE : 42764.78041172761
RMSE : 206.7964709847042
R2 : 0.895491990077968
Adjusted R2 :  0.8947224879340414

================Evalution Matrix=========================

    =========================================================    Evalution Graph    ===================================================
In [ ]:
#getting best estimator of xgb model given by xgb grid model
optimal_xgb_model=xgb_grid.best_estimator_
In [ ]:
#getting best param of rf model given by xgb_grid model
get_features_importance(optimal_xgb_model,X_train)
In [ ]:
#CatBoost
#installing catboost
!pip install catboost

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting catboost
  Downloading catboost-1.1.1-cp39-none-manylinux1_x86_64.whl (76.6 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 76.6/76.6 MB 10.8 MB/s eta 0:00:00
Requirement already satisfied: six in /usr/local/lib/python3.9/dist-packages (from catboost) (1.16.0)
Requirement already satisfied: matplotlib in /usr/local/lib/python3.9/dist-packages (from catboost) (3.7.1)
Requirement already satisfied: pandas>=0.24.0 in /usr/local/lib/python3.9/dist-packages (from catboost) (1.4.4)
Requirement already satisfied: graphviz in /usr/local/lib/python3.9/dist-packages (from catboost) (0.20.1)
Requirement already satisfied: plotly in /usr/local/lib/python3.9/dist-packages (from catboost) (5.13.1)
Requirement already satisfied: scipy in /usr/local/lib/python3.9/dist-packages (from catboost) (1.10.1)
Requirement already satisfied: numpy>=1.16.0 in /usr/local/lib/python3.9/dist-packages (from catboost) (1.22.4)
Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.9/dist-packages (from pandas>=0.24.0->catboost) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.9/dist-packages (from pandas>=0.24.0->catboost) (2022.7.1)
Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (1.0.7)
Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (4.39.3)
Requirement already satisfied: importlib-resources>=3.2.0 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (5.12.0)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (1.4.4)
Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (3.0.9)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (23.0)
Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (8.4.0)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.9/dist-packages (from matplotlib->catboost) (0.11.0)
Requirement already satisfied: tenacity>=6.2.0 in /usr/local/lib/python3.9/dist-packages (from plotly->catboost) (8.2.2)
Requirement already satisfied: zipp>=3.1.0 in /usr/local/lib/python3.9/dist-packages (from importlib-resources>=3.2.0->matplotlib->catboost) (3.15.0)
Installing collected packages: catboost
Successfully installed catboost-1.1.1
In [ ]:
#importing cat boost Regressor
from catboost import CatBoostRegressor
In [ ]:
# creating Catboost model
cb_model=CatBoostRegressor(verbose=0)
In [ ]:
#calling run_model to train,fit and evalution of catboost model
run_model(cb_model,X_train,y_train,X_test,y_test)

================Evalution Matrix=========================

MSE : 36706.5353729677
RMSE : 191.58949703198164
R2 : 0.910297049908164
Adjusted R2 :  0.9096365587892181

================Evalution Matrix=========================

    =========================================================    Evalution Graph    ===================================================
In [ ]:
#checking params of cat boost model
cb_model.get_all_params()
Out[ ]:
{'nan_mode': 'Min',
 'eval_metric': 'RMSE',
 'iterations': 1000,
 'sampling_frequency': 'PerTree',
 'leaf_estimation_method': 'Newton',
 'grow_policy': 'SymmetricTree',
 'penalties_coefficient': 1,
 'boosting_type': 'Plain',
 'model_shrink_mode': 'Constant',
 'feature_border_type': 'GreedyLogSum',
 'bayesian_matrix_reg': 0.10000000149011612,
 'eval_fraction': 0,
 'force_unit_auto_pair_weights': False,
 'l2_leaf_reg': 3,
 'random_strength': 1,
 'rsm': 1,
 'boost_from_average': True,
 'model_size_reg': 0.5,
 'pool_metainfo_options': {'tags': {}},
 'subsample': 0.800000011920929,
 'use_best_model': False,
 'random_seed': 0,
 'depth': 6,
 'posterior_sampling': False,
 'border_count': 254,
 'classes_count': 0,
 'auto_class_weights': 'None',
 'sparse_features_conflict_fraction': 0,
 'leaf_estimation_backtracking': 'AnyImprovement',
 'best_model_min_trees': 1,
 'model_shrink_rate': 0,
 'min_data_in_leaf': 1,
 'loss_function': 'RMSE',
 'learning_rate': 0.05512600019574165,
 'score_function': 'Cosine',
 'task_type': 'CPU',
 'leaf_estimation_iterations': 1,
 'bootstrap_type': 'MVS',
 'max_leaves': 64}
In [ ]:
#getting and visualizing features importance of catboost model
get_features_importance(cb_model,X_train)
In [ ]:
#lightGBM
import lightgbm as lgb
In [ ]:
lgbr=lgb.LGBMRegressor()
In [ ]:
# finding the best parameters for XGBRegressor by gridsearchcv
lgbr_para={'n_estimators': [150,200,250],'max_depth': [7,10,13]}
lgbr_grid=GridSearchCV(estimator=lgbr,param_grid=lgbr_para,cv=5,scoring='neg_mean_squared_error',verbose=5,n_jobs=-1)
In [ ]:
#calling run_model to train,fit and evalution of catboost model
run_model(lgbr_grid,X_train,y_train,X_test,y_test)

Fitting 5 folds for each of 9 candidates, totalling 45 fits

================Evalution Matrix=========================

MSE : 35410.75375394222
RMSE : 188.17745283094416
R2 : 0.9134636640470446
Adjusted R2 :  0.9128264890009115

================Evalution Matrix=========================

    =========================================================    Evalution Graph    ===================================================
In [ ]:
#getting and visualizing features importance of catboost model
get_features_importance(lgbr_grid.best_estimator_,X_train)
In [ ]:
#Combined Evaluation Matrix of All the models
In [ ]:
#creating dictionary to store all the metrices 
all_model_matrices={'Mean_square_error':mean_sq_error,'Root_Mean_square_error':root_mean_sq_error,'R2':r2_list,'Adjusted_R2':adj_r2_list}
In [ ]:
#list of all model made ,'Gradient_Boosting',,'CATBoost'
model_name=['Linear','Lasso','Ridge','Elasticnet','Polynomial','K-Nearyest_Neighbor','Decision_Tree','Random_Forest','Gradient_Boosting','Xtreme_GB','CATBoost','lightGBM']
In [ ]:
#converting dictionary to dataframe fro easy visual
matrices_df=pd.DataFrame.from_dict(all_model_matrices,orient="index",columns=model_name)
In [ ]:
matrices_df
Out[ ]:
Linear Lasso Ridge Elasticnet Polynomial K-Nearyest_Neighbor Decision_Tree Random_Forest Gradient_Boosting Xtreme_GB CATBoost lightGBM
Mean_square_error 175590.552873 175560.907118 175248.935066 175346.867499 118205.139666 83411.759209 90646.613699 62691.750320 54614.730269 42764.780412 36706.535373 35410.753754
Root_Mean_square_error 419.035264 418.999889 418.627442 418.744394 343.809743 288.810940 301.075761 250.383207 233.697947 206.796471 191.589497 188.177453
R2 0.572911 0.572983 0.573742 0.573504 0.712489 0.796159 0.778479 0.846795 0.866533 0.895492 0.910297 0.913464
Adjusted_R2 0.569766 0.569839 0.570603 0.570363 0.710372 0.794659 0.776848 0.845667 0.865551 0.894722 0.909637 0.912826
In [ ]:
#transposing matrices_df for easy and clear view and saving that as new final matrices df
final_matrices=matrices_df.transpose().reset_index().rename(columns={'index':'Models'})
In [ ]:
final_matrices
Out[ ]:
Models Mean_square_error Root_Mean_square_error R2 Adjusted_R2
0 Linear 175590.552873 419.035264 0.572911 0.569766
1 Lasso 175560.907118 418.999889 0.572983 0.569839
2 Ridge 175248.935066 418.627442 0.573742 0.570603
3 Elasticnet 175346.867499 418.744394 0.573504 0.570363
4 Polynomial 118205.139666 343.809743 0.712489 0.710372
5 K-Nearyest_Neighbor 83411.759209 288.810940 0.796159 0.794659
6 Decision_Tree 90646.613699 301.075761 0.778479 0.776848
7 Random_Forest 62691.750320 250.383207 0.846795 0.845667
8 Gradient_Boosting 54614.730269 233.697947 0.866533 0.865551
9 Xtreme_GB 42764.780412 206.796471 0.895492 0.894722
10 CATBoost 36706.535373 191.589497 0.910297 0.909637
11 lightGBM 35410.753754 188.177453 0.913464 0.912826
In [ ]:
#creating bar plot to visualize the Adjusted R2 of model 
plt.figure(figsize=(14,4),dpi=100)
sns.barplot(x='Models',y='Adjusted_R2',data=final_matrices.sort_values('Adjusted_R2'),palette="flare").set(title='Adjustted_R2 of Models')
plt.xticks(rotation=90);
In [ ]:
In [ ]:
#Model Explainibility
!pip install shap

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting shap
  Downloading shap-0.41.0-cp39-cp39-manylinux_2_12_x86_64.manylinux2010_x86_64.whl (572 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 572.4/572.4 KB 10.1 MB/s eta 0:00:00
Requirement already satisfied: scipy in /usr/local/lib/python3.9/dist-packages (from shap) (1.10.1)
Requirement already satisfied: scikit-learn in /usr/local/lib/python3.9/dist-packages (from shap) (1.2.2)
Collecting slicer==0.0.7
  Downloading slicer-0.0.7-py3-none-any.whl (14 kB)
Requirement already satisfied: packaging>20.9 in /usr/local/lib/python3.9/dist-packages (from shap) (23.0)
Requirement already satisfied: cloudpickle in /usr/local/lib/python3.9/dist-packages (from shap) (2.2.1)
Requirement already satisfied: numba in /usr/local/lib/python3.9/dist-packages (from shap) (0.56.4)
Requirement already satisfied: tqdm>4.25.0 in /usr/local/lib/python3.9/dist-packages (from shap) (4.65.0)
Requirement already satisfied: numpy in /usr/local/lib/python3.9/dist-packages (from shap) (1.22.4)
Requirement already satisfied: pandas in /usr/local/lib/python3.9/dist-packages (from shap) (1.4.4)
Requirement already satisfied: setuptools in /usr/local/lib/python3.9/dist-packages (from numba->shap) (67.6.1)
Requirement already satisfied: llvmlite<0.40,>=0.39.0dev0 in /usr/local/lib/python3.9/dist-packages (from numba->shap) (0.39.1)
Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.9/dist-packages (from pandas->shap) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.9/dist-packages (from pandas->shap) (2022.7.1)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.9/dist-packages (from scikit-learn->shap) (3.1.0)
Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.9/dist-packages (from scikit-learn->shap) (1.1.1)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.9/dist-packages (from python-dateutil>=2.8.1->pandas->shap) (1.16.0)
Installing collected packages: slicer, shap
Successfully installed shap-0.41.0 slicer-0.0.7
In [ ]:
from sklearn.tree import export_graphviz
import shap 
sns.set_style('darkgrid')
In [ ]:
for_instance=2
In [ ]:
# shape for xgboost model
# Initialize JavaScript visualizations in notebook environment
shap.initjs()
# Define a tree explainer for the built model
explainer = shap.TreeExplainer(xgb_grid.best_estimator_)
# obtain shap values for the first row of the test data
shap_values = explainer(X_test,check_additivity=False)
# # visualize the first prediction's explanation with a force plot
shap.plots.force(shap_values[for_instance])
Out[ ]:
Visualization omitted, Javascript library not loaded!
Have you run `initjs()` in this notebook? If this notebook was from another user you must also trust this notebook (File -> Trust notebook). If you are viewing this notebook on github the Javascript has been stripped for security. If you are using JupyterLab this error is because a JupyterLab extension has not yet been written.
In [ ]:
#shape for catboost model
# Initialize JavaScript visualizations in notebook environment
shap.initjs()
# Define a tree explainer for the built model
explainer = shap.TreeExplainer(cb_model)
# obtain shap values for the first row of the test data
shap_values = explainer(X_test,check_additivity=False)
# # visualize the first prediction's explanation with a force plot
shap.plots.force(shap_values[for_instance])
Out[ ]:
Visualization omitted, Javascript library not loaded!
Have you run `initjs()` in this notebook? If this notebook was from another user you must also trust this notebook (File -> Trust notebook). If you are viewing this notebook on github the Javascript has been stripped for security. If you are using JupyterLab this error is because a JupyterLab extension has not yet been written.
In [ ]:
#shap for lightGBM model
# Initialize JavaScript visualizations in notebook environment
shap.initjs()
# Define a tree explainer for the built model
explainer = shap.TreeExplainer(lgbr_grid.best_estimator_)
# obtain shap values for the first row of the test data
shap_values = explainer(X_test,check_additivity=False)
# # visualize the first prediction's explanation with a force plot
shap.plots.force(shap_values[for_instance])
Out[ ]:
Visualization omitted, Javascript library not loaded!
Have you run `initjs()` in this notebook? If this notebook was from another user you must also trust this notebook (File -> Trust notebook). If you are viewing this notebook on github the Javascript has been stripped for security. If you are using JupyterLab this error is because a JupyterLab extension has not yet been written.
In [ ]:
#ELI5
!pip install eli5

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting eli5
  Downloading eli5-0.13.0.tar.gz (216 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 216.2/216.2 KB 5.2 MB/s eta 0:00:00
  Preparing metadata (setup.py) ... done
Requirement already satisfied: attrs>17.1.0 in /usr/local/lib/python3.9/dist-packages (from eli5) (22.2.0)
Requirement already satisfied: jinja2>=3.0.0 in /usr/local/lib/python3.9/dist-packages (from eli5) (3.1.2)
Requirement already satisfied: numpy>=1.9.0 in /usr/local/lib/python3.9/dist-packages (from eli5) (1.22.4)
Requirement already satisfied: scipy in /usr/local/lib/python3.9/dist-packages (from eli5) (1.10.1)
Requirement already satisfied: six in /usr/local/lib/python3.9/dist-packages (from eli5) (1.16.0)
Requirement already satisfied: scikit-learn>=0.20 in /usr/local/lib/python3.9/dist-packages (from eli5) (1.2.2)
Requirement already satisfied: graphviz in /usr/local/lib/python3.9/dist-packages (from eli5) (0.20.1)
Requirement already satisfied: tabulate>=0.7.7 in /usr/local/lib/python3.9/dist-packages (from eli5) (0.8.10)
Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.9/dist-packages (from jinja2>=3.0.0->eli5) (2.1.2)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.9/dist-packages (from scikit-learn>=0.20->eli5) (3.1.0)
Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.9/dist-packages (from scikit-learn>=0.20->eli5) (1.1.1)
Building wheels for collected packages: eli5
  Building wheel for eli5 (setup.py) ... done
  Created wheel for eli5: filename=eli5-0.13.0-py2.py3-none-any.whl size=107747 sha256=e95dd78dd376d4cd473743aeff3e054d24e839f78cb7ca97fcb36daffbb83d2c
  Stored in directory: /root/.cache/pip/wheels/7b/26/a5/8460416695a992a2966b41caa5338e5e7fcea98c9d032d055c
Successfully built eli5
Installing collected packages: eli5
Successfully installed eli5-0.13.0
In [ ]:
import eli5 as eli
feature_names=list(X_test.columns)
for_instance=2
In [ ]:
#eli5 for xgboost model 
eli.show_prediction(xgb_grid.best_estimator_, X_test.reset_index(drop=True).loc[[for_instance]],
                    feature_names=feature_names,
                    show_feature_values=True)
Out[ ]:

y (score 694.064) top features

Contribution? Feature Value
+705.296 <BIAS> 1.000
+225.114 Solar_Radiation 1.680
+124.098 Temperature 34.000
+55.327 Humidity 50.000
+54.859 Functioning_Day_Yes 1.000
+15.271 weekdays_weekend 1.000
+9.990 month 7.000
+0.447 Holiday_No Holiday 1.000
+0.011 Seasons_Summer 1.000
+0.002 Seasons_Spring 0.000
+0.000 Snowfall 0.000
-0.107 Rainfall 0.000
-0.500 Visibility 1744.000
-13.444 Wind_speed 1.200
-164.349 Dew_point_temperature 22.100
-317.950 Hour 14.000
In [ ]:
#eli5 for lgbr model 
eli.show_prediction(lgbr_grid.best_estimator_, X_test.reset_index(drop=True).loc[[for_instance]],
                    feature_names=feature_names,
                    show_feature_values=True)
Out[ ]:

y (score 710.232) top features

Contribution? Feature Value
+705.796 <BIAS> 1.000
+191.367 Solar_Radiation 1.680
+74.526 Temperature 34.000
+38.931 Functioning_Day_Yes 1.000
+37.236 Humidity 50.000
+27.307 weekdays_weekend 1.000
+12.110 Seasons_Spring 0.000
+10.303 Seasons_Summer 1.000
+10.040 Wind_speed 1.200
+7.519 Rainfall 0.000
+3.179 month 7.000
+2.952 Holiday_No Holiday 1.000
+2.872 Visibility 1744.000
+2.695 Seasons_Winter 0.000
+1.149 Snowfall 0.000
-119.583 Dew_point_temperature 22.100
-298.167 Hour 14.000